Subnet calculations....

[ProPlus]

Friend of mine needs help with the following.... Im not a network bod only a code monkey so I figured someone here might help....

Consider the IP scenarios below and fill in the missing details in the table. You are only required to fill in the details for the first 5 usable subnets. Complete the following table of information for each question. Make sure that you use the correct rules for each set of subnet calculations

Host Bits:
Subnetwork Bits
Number of Usable Hosts:
Number of Usable Subnets:
New Subnet Mask:

Number | Network Add | Broadcast Add | First Host Add | Last Host Add|
1
2
3
4
5

Part A: Class Based Subnet Calculations
1) Complete the table for the Class C network 196.78.1.0, and create a subnet scheme containing at least four subnets and maximising the number of hosts on each networks
2) Complete the table for the Class B network 130.80.0.0, and create a subnet scheme where there are at least 300 hosts on each network maximising the number of available networks.
Part B: CIDR/VLSM Based Subnet Calculations
1) Complete the table for the network 195.95.4.0/22, and create a subnet scheme containing 8 subnets and maximising the number of hosts on each network.
2) Complete the table for the network 140.40.64.0/18, and create a subnet scheme where there are at least 500 hosts on each network, maximising the number of available networks.

[plasticpig]

www.subnet-calculator.com is your freind

[guydw]

This is all very easy.....

Quick overview...

An IP address has 32 bits (IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII) divided into four lots of eight bits (an octet) - 11111111.11111111.11111111.11111111

If you make the right most bit in each octet 1, then double each one as you move left, the value when it is all ones is 255 (so we have 255.255.255.255). So if you include zero you have 256 possible values.. 00000001 = 1, 00000010 =2, 00000011=3, 00000100=4 and so on).

Nowadays it is incorrect to talk about classes, as it is out dated (I won't go into why, but you have something called Classless interdomain routing) - but basically a class C is used to mean an address where you have 256 option i.e. the first three octets are fixed, and you can change the last one, a class B is one where the first two octets is fixed...

However in practice you would never use a class B on your network as you would waste many possible IP addresses, so you divide it up. Here's where it gets fun - you use a subnet mask... basically this tells you what part of your IP address is for the network (so stays the same) and which part is for the host (changes). Think of this as being like street numbers - the name of the street stays the same (the network address) and the house number changes, and is uniquely allocated to the house (host address).

So if the bits on your subnet mask (32 bits, in four octests, same as the IP address) is a one, then that part of the IP address is part of the network portion of the IP address, if it is a zero it is part of the host address.... you with me?

so network: 11000000.10101000.00000001.00000000
mask: 11111111.11111111.11111111.00000000

gives you 192.168.1.0, mask 255.255.255.0 - you will notice that the first 24 bits are a 1, that is they define that the first 24 bits of the IP address are the network part, or the first three octets, this is a class C and is written 192.168.1.0/24, called a slash 24.

Now, this network, or subnetwork, has a number of fixed values ...

The network number (in this case .0, so 192.168.1.0) it is always the first number in the range...

the broadcast address, always the last number, in this case 255.

The first USEABLE address i.e. it can be assigned to a host, the number after the network address, in this case .1

The last USEABLE address, the number before the broadcast address, in this case 254.

In reality before you can assign addresses to users you give an address to your router, generally .1 (this would be the default gateway to PCs on your network) and sometimes other addresses are allocated for network equipment.

So to "subnet" a Class C, you divide it up using the bits from the subnet mask ....

so (we'll assume 192.168.1.0) looking at the last octet you have....

a subnet mask of :

10000000 = 128,
this gives you 128 host addresses (256 - 128)-remember I included the zero so that is two subnets of a class C ....
network addresses of .0, and .128, broadcast addresses of .127 and .255, useable ranges of 1-126 and 129 to 254.

then you have 11000000, gives you 64 hosts (just add up the value of the last 6 bits, or easier, take the subnet mask, which is 192, and subtract that from 256, gives you 64..)

So, network addresses, .0, .64, .128, .192, broadcast .63,.127,.191,.255

One thing to note is using subnet zero - if you use this some of the maths doesn't work, but in practice the wasteful behaviour of not using subnet zero was discarded years ago.

so...



Part A: Class Based Subnet Calculations
1) Complete the table for the Class C network 196.78.1.0, and create a subnet scheme containing at least four subnets and maximising the number of hosts on each networks

Just take 256 and divide it by four, you get 64 hosts, 192 mask (/26)

2) Complete the table for the Class B network 130.80.0.0, and create a subnet scheme where there are at least 300 hosts on each network maximising the number of available networks.

again, just do the maths, must be bigger than a /24, so a /23 is the next one, giving us 512 hosts

Part B: CIDR/VLSM Based Subnet Calculations
1) Complete the table for the network 195.95.4.0/22, and create a subnet scheme containing 8 subnets and maximising the number of hosts on each network.

this means you have FOUR class C's available, so to divide it by eight you chop your Class C's (/24) in half, giving you eight /25's each with 128 ...

2) Complete the table for the network 140.40.64.0/18, and create a subnet scheme where there are at least 500 hosts on each network, maximising the number of available networks.

at least 500 - we want a bunch of /23's (512), /18 is 64 class Cs or /24s, so it gives you 32 /23's

Number | Network Add | Broadcast Add | First Host Add | Last Host Add|
1
2
3
4
5

[guydw]

blimey, I just looked at that.....

[victormeldrew]

Oh yes, all VERY easy.

[guydw]

yeah fair point......

to be honest though, if you are a network bod and can't just do this stuff in your head, you've got no chance.....

[agent006]

Subnetting is a fine example of a concept that is actually really quite simple, but almost impossible to explain simply.

[Zad]

Hehe, I promise, it is easy when your brain "gets" it. It does look fearsomely complex though.

[agent006]

I usually try and compare networks to roads when trying to explain stuff. Doesn't really work with subnetting though.

[guydw]

Yeah I agree, but I'm sure that if you read back what I wrote, and spend 5 mins going through it bit by bit it makes sense.

People that don't really understand how IP addressing works, or how to subnet in their head really annoys me (if they claim to be in networking...).

There's this mad guy called Heinz who runs a CCIE bootcamp, and he constantly gives you an IP address and a mask, and you have 2 secs to tell him if it is a valid address, the network address and the broadcast - if you can't do this then you have no chance ... and if you don't just know this stuff you won't spot incorrectly addressed interfaces and so on.